Electromagnetic Distance-Sensing System

This application claims the benefit of European Patent Application No. 24305525.8 filed Apr. 4, 2024, the disclosure of which is incorporated herein by reference in its entirety.

This disclosure relates to electromagnetic distance sensors, including, but not limited to, eddy-current sensors.

Electromagnetic sensors can be used for variety of contactless sensing applications including proximity sensing of conductive materials. Some sensors can be used to output a quantitative measure of distance, rather than just binary proximity sensing.

It can be desirable to place multiple electromagnetic sensor coils close to one another to form a multi-sensor distance-sensing system. This may allow a sensing system to gather more information about the location of a conductive target and increase system reliability. However, where multiple electromagnetic sensor coils are in close proximity to one another, mutual inductance between the respective sensors can cause interference that can impair the performance of the distance sensing.

A multi-sensor electromagnetic sensor system is desired that is robust against such interference.

From a first aspect, the disclosure provides an electromagnetic sensor system for sensing distance to a conductive target. The sensor system includes: a first sensor comprising: a first electromagnetic sensor coil; first drive circuitry for driving the first electromagnetic sensor coil with a first alternating current at a first predetermined frequency; and first sensing circuitry, electrically coupled to the first electromagnetic sensor coil, and configured to output a first signal having a component that is indicative of a distance from the first electromagnetic sensor coil to the conductive target. The system also includes a second sensor comprising: a second electromagnetic sensor coil; and second drive circuitry for driving the second electromagnetic sensor coil with a second alternating current at a second predetermined frequency, wherein the second predetermined frequency is offset from the first predetermined frequency by an offset amount. The system also includes filtering circuitry configured to apply a low-pass filter to the first signal, wherein the low-pass filter has a cut-off frequency that is below the offset amount such that a component of the first signal caused by mutual inductance between the first and second electromagnetic sensor coils is attenuated, and to output a first filtered signal indicative of the distance from the first electromagnetic sensor coil to the conductive target.

From a second aspect, the disclosure provides a method for sensing distance to a conductive target using an electromagnetic sensor system. The sensor system includes: a first sensor comprising a first electromagnetic sensor coil; and a second sensor comprising a second electromagnetic sensor coil. The method includes: driving the first electromagnetic sensor coil with a first alternating current at a first predetermined frequency; driving the second electromagnetic sensor coil with a second alternating current at a second predetermined frequency, wherein the second predetermined frequency is offset from the first predetermined frequency by an offset amount; outputting a first signal having a component that is indicative of a distance from the first electromagnetic sensor coil to the conductive target; applying a low-pass filter to the first signal, wherein the low-pass filter has a cut-off frequency that is below the offset amount such that a component of the first signal caused by mutual inductance between the first and second electromagnetic sensor coils is attenuated; and outputting a first filtered signal indicative of the distance from first electromagnetic sensor coil to the conductive target.

When the electromagnetic sensor coils are arranged close enough together, the magnetic field generated by each sensor coil has an inductive effect on the other electromagnetic sensor coil. Thus, for the first electromagnetic sensor, interference in the output of the first sensing circuitry may occur as a result of mutual inductance between the first and second electromagnetic sensor coils. In examples in which the second sensor also has sensing circuitry, interference may also be present in its output.

The effect of this interference in the output may be periodic, having a frequency equal to the difference between the driving frequency of the first electromagnetic sensor coil and the driving frequency of the second electromagnetic sensor coil. Thus it will be seen that, by driving the first electromagnetic sensor coil at a first predetermined frequency which is offset by a predetermined offset amount from the second predetermined frequency (i.e. higher than the second predetermined frequency by the predetermined offset amount or lower than the second predetermined frequency by the predetermined offset amount, the offset amount therefore being an absolute value of the offset), the low-pass filter may be designed to attenuate frequencies in the first signal that are equal to and greater than the offset amount, and so mitigate the effect of interference resulting from mutual inductance between the two electromagnetic coils. The stability and/or accuracy of the signal indicative of the distance from the first electromagnetic sensor coil to the conductive target may therefore be improved.

In some examples the first sensor is an inductive sensor, e.g. an eddy-current sensor. The first electromagnetic sensor coil may comprise a ferromagnetic core, but in some examples does not comprise a ferromagnetic core—e.g. the first electromagnetic sensor coil may comprise an air core. Not using a ferromagnetic core may reduce magnetic losses and thermal nonlinearities in the sensor output. This may improve the responsiveness of the sensor output to changes in distance to the conductive target. However, using a ferromagnetic core may focus the magnetic field produced by the first electromagnetic sensor coil, and thus allow the first sensor to detect the conductive target from further away. The first electromagnetic sensor coil may comprise one or more turns of conductive wire. Increasing the number of turns of conductive wire may increase the strength of the magnetic field produced by the first electromagnetic sensor coil. A stronger magnetic field may increase the sensing range of the first sensor. On the other hand, decreasing the number of turns of conductive wire may increase the weight of the electromagnetic sensor system. Weight may be a relevant design consideration in, for example, aircraft systems.

The component of the first signal indicative of distance may be a low-frequency component. It may be a constant or slowly changing DC component. It may have a maximum frequency that is less than the offset amount, e.g. less than a half or a tenth of the offset amount. It may have a maximum frequency that is less than the first predetermined frequency and less than the second predetermined frequency. In some examples the component of the first signal may be representative of an impedance of the first electromagnetic sensor coil. The impedance of the first electromagnetic sensor coil may vary depending on a distance (e.g. an air gap) between the conductive target and the first electromagnetic sensor coil—e.g. the impedance of the first electromagnetic sensor coil may decrease as the distance decreases.

In some examples, the first sensor (e.g. the first sensing circuitry) may comprise a Wheatstone bridge. The first electromagnetic sensor coil may form at least part of one branch of the Wheatstone bridge. A further branch of the Wheatstone bridge may comprise a reference coil. The Wheatstone bridge may be configured to produce a differential signal representative of the impedance of the first electromagnetic sensor coil. The differential signal may be a differential current, but in some examples is a differential voltage.

In some examples, the first sensor (e.g. the first sensing circuitry) comprises first conditioning circuitry which is configured to receive the differential signal from the Wheatstone bridge, and output a single-ended output signal. The single-ended output signal may, in some examples, be a time-averaged (e.g. mean) value. The first conditioning circuitry may comprise analogue and/or digital circuitry for performing any one or more of: amplification, impedance adaptation, rectification, differential-to-single-ended conversion, or time averaging. The first conditioning circuitry may be arranged to output the first signal to the filtering circuitry.

In some examples, the offset amount is equal to or greater than 1 kHz, e.g. equal to or greater than 10 kHz. This should be understood to mean that the second predetermined frequency used to drive the second electromagnetic sensor coil is at least 1 kHz higher than or at least 1 kHz lower than the first predetermined frequency used to drive the first electromagnetic sensor coil. In a further example, the second predetermined frequency used to drive the second electromagnetic sensor coil is at least 10 kHz higher than or at least 10 kHz lower than the first predetermined frequency used to drive the first electromagnetic sensor coil. A greater offset amount may allow the cut-off frequency of the low-pass filter to be increased. Maximising the offset amount within the constraints of the electromagnetic sensor system (e.g. subject to maximum and/or minimum frequencies supported by the drive circuitry) may increase the range of frequencies which are passed through by the low-pass filter—i.e. the bandwidth of the output signal provided by the electromagnetic sensor system may be increased. An output signal with greater bandwidth may allow the system to be more responsive to changes in the distance to the conductive target over time.

The cut-off frequency may be the −3 dB frequency of the low-pass filter. The low-pass filter may be configured to attenuate the mutual-inductance component by at least 3 dB or at least 6 dB or at least 50 dB. In some examples, the cut-off frequency of the low-pass filter is one-tenth, or one-hundredth, of the offset amount or lower. Implementing a cut-off frequency which is at most one-tenth of the offset amount may result in the signal component of the output of the first sensor at a frequency corresponding to the offset amount being attenuated further, e.g. by at least 50 dB. The low-pass filtering circuitry may comprise any circuitry suitable for implementing a low-pass filter—e.g. analogue circuitry and/or digital circuitry. The low-pass filter may be a digital or analogue filter. It may comprise active circuitry or fast-Fourier-transform (FFT) circuitry.

The filtered signal may decrease with increasing distance between the first electromagnetic sensor coil and the conductive target. It may decrease linearly in some examples, although this is not essential. It may be an analogue signal—e.g. a voltage signal. The electromagnetic sensor system may comprise an analogue to digital converter (ADC). It may be arranged to output the first filtered signal as a first digital signal. It may comprise a processing system (e.g. comprising an ASIC, or a DSP, or a microcontroller) configured to process the first digital signal. A predetermined relationship between the output of the first sensor and the distance between the conductive target and the first sensor may be stored in a memory of the electromagnetic sensor system, e.g. as a look-up table in a database or as a mathematical relationship, and the predetermined relationship may be used by a processing system (in hardware or by software) to calculate a calibrated distance between the conductive target and the first electromagnetic sensor coil (e.g. in millimetres). The sensor system may generate a time series of digital distance values.

In other, potentially overlapping examples, the signal indicative of the distance to the conductive target may be used to determine a binary state—e.g. whether or not the conductive target is present within a sensing range of the first electromagnetic sensor coil, and/or whether or not the conductive target is closer than a predetermined threshold value to the first electromagnetic sensor coil. In such examples, the signal indicative of the distance to the conductive target may be input to a comparator, and the output from the comparator may indicate whether the conductive target is closer to the first electromagnetic sensor than a threshold value. Thus, it can be seen that the electromagnetic sensor system may provide a proximity detection system and/or a distance measurement system.

In some examples, each of the first sensor and the second sensor may be configured to sense a respective distance to the same conductive target. The conductive target may, in some examples, be a component of the electromagnetic sensor system. In some examples, the first sensor coil and the second sensor coil may be arranged or mounted such that a coil axis of the first sensor is parallel to a coil axis of the second sensor. The coils may be arranged or mounted in a same plane. They may be separated by less than a hundred, or ten, or five times the diameter of the first coil.

In some examples the second sensor is an inductive sensor, e.g. an eddy-current sensor. The second sensor may comprise second sensing circuitry, electrically coupled to the second electromagnetic sensor coil, and configured to output a second signal having a component that is indicative of a distance from the second electromagnetic sensor coil to the, or a further, conductive target. The filtering circuitry may be further configured to apply a low-pass filter to the second signal, wherein the low-pass filter has a cut-off frequency that is below the offset amount such that a component of the second signal caused by mutual inductance between the first and second electromagnetic sensor coils is attenuated, and to output a second filtered signal indicative of the distance from the second electromagnetic sensor coil to the conductive target. The same low-pass filter, or a further identical low-pass filter, may be applied to the second signal as to the first signal. The second sensing circuitry may comprise second conditioning circuitry that may be identical to, or shared with, or may have features in common with, the first conditioning circuitry. The second sensor may have any of the features of the first sensor as disclosed herein. The sensor system may be configured to output the second filtered signal as a digital signal.

Using a plurality of sensors to sense distance to the same conductive target allows two independent signals indicative of distance to the conductive target to be generated. This may improve the reliability of the sensing system. Some examples where a respective distance from each electromagnetic sensor coil to the conductive target is determined may be further configured to determine (e.g. calculate) a mean distance value from these two distances. In other examples where the signal indicative of the distance to the conductive target is used to indicate a binary state—e.g. the conductive target being present or not present, one output signal may be used to corroborate the output signal of the other. In this example, a resulting output signal indicating the presence of the conductive target may only be set to a positive indication if both intermediate output signals indicate that the conductive target is present. Furthermore, having at least two different sensors in the electromagnetic sensor system may add redundancy to the system, and thus improve its safety and reliability.

In some examples the electromagnetic sensor system comprises more than two sensors, each sensor comprising a respective electromagnetic sensor coil and drive circuitry for driving said electromagnetic sensor coil at a respective predetermined frequency. The filtering circuitry may, in said examples, be configured to apply a low-pass filter to a respective signal output by at least one of the plurality of sensors, wherein the low-pass filter has a cut-off frequency that is below a minimum offset amount such that a component of the signal caused by mutual inductance between any of the electromagnetic sensor coils is attenuated. The minimum offset amount may correspond to (e.g. equal) the smallest offset amount between any two of the predetermined frequencies at which the respective electromagnetic sensor coils in the electromagnetic sensor system are driven. As such, the low-pass filter is designed to attenuate signal components at each of the frequencies corresponding to the respective predetermined offset amounts between any combination of predetermined driving frequencies used in the electromagnetic sensor system.

Any features of the first sensor described in the examples herein may be applicable to any further sensors in the electromagnetic sensor system.

In some examples, the electromagnetic sensor system may be used as part of a sensor apparatus of a propeller blade angle detection system. In the propeller blade angle detection system, the conductive target may be elongate and integral with, or configured for coupling to, a component of a variable pitch propeller assembly so as to rotate coaxially with the propeller about an axis of rotation of the propeller, and to translate axially with varying pitch angle of one or more blades of the propeller. The conductive target may comprise an outer surface that is a surface of revolution about the axis of rotation, and that has a diameter that changes monotonically with distance along the axis of rotation. Each of the first sensor and the second sensor may be configured for non-rotational and non-translating mounting adjacent the conductive target, and each of the first sensor and the second sensor may be configured to output a respective signal measuring a respective distance between the outer surface of the conductive target and the respective electromagnetic sensor coil. The sensor apparatus may be configured to output a signal in dependence on either or both of the measured distances, wherein the output signal is indicative of the pitch angle of the one or more blades of the propeller.

Features of any example described herein may, wherever appropriate, be applied to any other example described herein. Where reference is made to different examples or sets of examples, it should be understood that these are not necessarily distinct but may overlap.

Electromagnetic sensors can be used for variety of contactless sensing applications including fault detection in conductive materials, as well as qualitative (e.g. binary) proximity or quantitative distance sensing of conductive materials. An eddy-current sensor comprises an electromagnetic sensor coil which is driven by an oscillator. Driving the sensor coil with an alternating current produces a time-varying magnetic field. When a conductive target passes through this magnetic field, according to Faraday's law of induction, the magnetic field induces alternating electric currents in the surface of the conductive target. These induced surface currents are known as eddy currents. The eddy currents self-produce another magnetic field, which affects the impedance of the sensor coil. The change in impedance of the sensor coil due to the eddy currents in the conductive target is directly linked to the distance (e.g. air gap) between the sensor coil and the conductive target. Thus, the distance can be obtained by measuring the equivalent impedance. This allows eddy-current sensors to function as distance and proximity sensors. Having an array of sensors can allow more information regarding the proximity of different parts of the conductive target to be obtained—e.g. for improved measurement accuracy, or for determining additional information about the position or orientation of the conductive target.

Where multiple electromagnetic sensors are mounted in close proximity to one another, mutual inductance between the sensor coils in the respective sensors has an effect on the impedance of each sensor coil. If the driving frequencies of the different electromagnetic sensors are not the same, the mutual inductance generated in a first sensor coil due to the effects from a proximate second sensor coil varies periodically over time. This phenomenon is known as cross-talk between sensors, and can affect the accuracy of the sensor outputs. Examples of the present disclosure seek to address this issue.

is a schematic diagram showing a sensor arraycomprising two electromagnetic sensors,. In the example described herein each of the sensors,is an eddy-current proximity sensor, although in other examples the electromagnetic sensors,could be two different types of inductive (i.e. coil-based) sensor. Whilst there are two electromagnetic sensors shown in, the present disclosure may be applicable to arrays comprising more than two electromagnetic proximity sensors—e.g. arranged in a line or as a rectangular array. The coils of the sensors,may have parallel axes and may lie in a common plane, although this is not essential.

As shown in, the first sensoris arranged to output a first analogue output signal VOUT_1 to a processing system(e.g. comprising a microcontroller with a memory storing software for performing operations as described herein). The second sensoris arranged to output a second analogue output voltage VOUT_2 to the processing system. The processing systemmay be configured to process the output signals from the first sensorand second sensorand output information regarding the distance between a conductive target and the sensor array. Whilst the processing systeminis schematically shown as one block, receiving both output signals VOUT_1 and VOUT_2, it should be understood that processing systemmay encompass multiple analogue and/or digital circuit components which may be electrically and/or physically separate to one another. Each of the sensors,may contain or be coupled to a different respective oscillator. Driving each sensor using a separate alternating supply, rather than a single common supply, can add redundancy to the system and allow the systemstill to sense distance even if one of the oscillators fails. This may be important in safety-critical systems, where the failure of one component should not result in the failure of the whole system. This may apply to aircraft systems, for example.

is a schematic diagram showing exemplary circuitryfor implementing an eddy-current proximity sensor as implemented in each of the first sensorand the second sensorshown in. The circuitrycomprises a Wheatstone bridge having a sensor coilon a first branch, a reference coilon a second branch, a first balancing resistors Rson a third branch, and a second balancing resistor Rson a fourth branch. An oscillatorprovides an excitation voltage across the Wheatstone bridge to energize the sensor coil. A conductive targetis shown in close proximity to the sensor coil. As explained above, the impedance of the sensor coilwill change depending on the proximity of a conductive targetto the sensor coil. The sensor coilis therefore modelled as a variable inductor and variable resistor. The reference coilis modelled as an inductor with fixed inductance and a resistor with fixed resistance. A first sensing voltage Vsense_A and a second sensing voltage Vsense_B are output from the Wheatstone bridge to conditioning circuitry. The analogue conditioning circuitryis followed by filtering circuitrywhich together use the change in the sensing voltages to determine a sensor output signal VOUT_1 having a steady voltage level that is indicative of a distance from the respective sensor coil to the conductive target.

is a schematic diagram showing an exemplary implementation of the conditioning circuitryshown in. The conditioning circuitryreceives the first sensing voltage Vsense_A and a second sensing voltage Vsense_B from the Wheatstone bridge shown in. This differential signal is passed through, in order: a first impedance adaptation and gain stage; a full wave rectifier; a second impedance adaptation and gain stage; and a differential-to-single-ended conversion stage, which is configured to output a single-ended voltage signal representative of the proximity of the conductive targetto the sensor coil. This single-ended signal is input to a mean value extraction stage. Extracting a mean valuemay improve the stability of the signal which is output from the conditioning circuitry. The output of the mean value extraction stageis then passed to filtering circuitry, which is arranged to apply a low-pass filter as described herein. Each of these different stages of circuitry could be implemented in a number of different ways, and some examples may omit one or more of them, or include additional stages, or perform these operations in a different order. For example, the low-pass filterhere receives a single-ended input but could, in some examples, receive and filter a differential signal.

For each sensor,, any of all of the components and connections that form the Wheatstone bridge, apart from the sensor coilitself, as well as any or all of the components of the conditioning circuitry, may collectively implement the “sensing circuitry” as disclosed herein.

The outputs VOUT_1, VOUT_2 from the two sensors,may be further processed by the processing systemshown in. This systemmay comprise components for performing analogue signal processing and/or digital signal processing. It may include one or more analogue or digital displays for displaying information to a user. The processing systemmay include one or more analogue to digital converters (ADCs) for sampling the voltages VOUT. e.g. comprising each being compared against a voltage threshold by an analogue or digital comparator, or comprising each being converted into a digital value representative of a respective distance to the conductive targetfor further digital processing and/or display to a user.

is a plotshowing the form of an exemplary output signal VOUT_1 from the conditioning circuitryof the eddy-current proximity sensorshown inin one example use case, before the output signal is input to the filtering circuitry. The plotshows the output signal VOUT_1 from the conditioning circuitryover a first period of time T1 in which the conductive targetis located at a first distance from the sensor coil, and a later, second period of time T2, in which the conductive targetis located at a second distance from the sensor coil.

During the first period of time T1 the output signal VOUT_1 takes the approximate form of:

During the second period of time T2 the output signal Vtakes the approximate form of:

The sinusoidal component in the output signal VOUT_1 is caused by cross-talk interference from the second sensor. The frequency of this sinusoidal variation θ can be calculated as being equal to the difference between the driving frequency ω1 of the first coil, and the driving frequency ω2 of the second coil—i.e. θ=|ω−ω|, where the first sensor input signal is modelled as V=A sin ωt, and the second sensor input signal is modelled as V=B sin ωt.

A constant or slowly-changing DC (i.e. low-frequency) component in the output signal VOUT_1 varies with the separation distance between the first sensor coiland the conductive target. Thus, the signal component Vis representative of the separation distance during the first time period T1, and the signal component Vis representative of the separation distance during the second time period T2. The relationship between the output signal VOUT_1 and the separation distance may vary in dependence on the shape of the conductive target.

In the example shown in the plot, there is a significant difference between the driving frequency of the first sensorand the second sensor. The absolute value of ω−ω—i.e. the offset amount referred to above, can be chosen for a given application based on a required sensitivity in the measurement of distance to the targetover time. The frequency of the sinusoidal variation (θ=|ω−ω|) may be chosen to be high compared to the maximum expected variation in the output signal due to a change in separation between the sensor coil and the conductive target.

Purposefully selecting the driving frequency of the first coil and the driving frequency of the second coil so that they are substantially different to one another enables the relatively low-frequency component of the sensing signal which is of interest to be isolated using a low-pass filter. The relatively high frequency of the sinusoidal variation in the output signal is filtered out by a low-pass filter, where the low-pass filter is designed to attenuate the frequency of the sinusoidal variation caused by the expected interference effect. The greater the difference between the driving frequency of the first coil and the driving frequency of the second coil, the more bandwidth is available for the remaining sensing signal produced by the sensor system after the output signal of the conditioning circuitryis passed through the low-pass filter in the filtering circuitry. A greater bandwidth remaining in the output signal may allow the system to be more responsive to changes in the proximity of the conductive targetto the sensor coil.

is a plot showing an example action of a low-pass filter that may be used to attenuate the sinusoidal variation in the output signal VOUT_1 of the first sensorin some examples. A second similar or identical low-pass filter may be used to filter the output signal VOUT_2 of the second sensor.

The plot shown inshows, on the right side vertical axis, the gainof the low-pass filter in decibels (dB) as it varies with frequency. It also shows, on the left side vertical axis, the output frequency spectrumof the first sensor output signal VOUT_1. The output frequency spectrumpeaks at frequency ƒ(1100 kHz in this example), where

As can be seen in, the attenuation of the low-pass fileris around −92 dB at this peak, ƒ, in the output frequency spectrum. The low-pass filtermay be designed such that the-cut-off frequency ƒ(1 kHz in this example) is approximately a factor of ten less than the peak frequency of the output frequency spectrum. The horizontal linerepresents an attenuation objective with a crosstalk error of less than 0.1% full scale.

In examples where eddy-current proximity sensors are used in control systems which need to be highly responsive, a corresponding minimum bandwidth may be required to remain in the sensing system. For example, a minimum of 1100 Hz of bandwidth may be required to remain in the output signal from the conditioning circuitry. The difference between the driving frequency of the first sensor and the second sensor may therefore be set to be approximately 1 kHz or greater, such that the low-pass filtering circuitrydoes not substantially attenuate frequencies at 1100 Hz or less.

is a plotshowing an exemplary output signal from the low-pass filtershown inwhen it is applied to the output signal VOUT_1 of the eddy-current sensorshown in. As can be seen, the variation in the low-frequency (i.e. slowly varying DC) component of the output signal VOUT_1 (e.g. Vand V), which could be seen in, remains, but the high-frequency sinusoidal variation due to cross-talk from mutual inductance, which could previously be seen in, has been filtered out.

Whilst the example is described above with reference to mitigating cross-talk interference between only two sensors, the principles of the present disclosure can equally apply to arrays of more than two sensors located in close proximity with one another. Where there are more than two sensors, the cut-off frequency of the low-pass filtermay be designed such that it attenuates frequencies which are greater than or equal to the smallest difference in frequency between the frequencies of any two driving supplies—e.g. the filter may be designed such that

where ωand ωare driving frequencies used in the sensor array for different respective sensors m and n, and the cut-off frequency ƒis chosen to be a factor of ten lower than the expected minimum cross-talk frequency. By cutting-off from the smallest expected difference in frequency, all larger differences in frequencies will also be filtered out.